Research

Paper

TESTING February 23, 2026

In-context Pre-trained Time-Series Foundation Models adapt to Unseen Tasks

Authors

Shangqing Xu, Harshavardhan Kamarthi, Haoxin Liu, B. Aditya Prakash

Abstract

Time-series foundation models (TSFMs) have demonstrated strong generalization capabilities across diverse datasets and tasks. However, existing foundation models are typically pre-trained to enhance performance on specific tasks and often struggle to generalize to unseen tasks without fine-tuning. To address this limitation, we propose augmenting TSFMs with In-Context Learning (ICL) capabilities, enabling them to perform test-time inference by dynamically adapting to input-output relationships provided within the context. Our framework, In-Context Time-series Pre-training (ICTP), restructures the original pre-training data to equip the backbone TSFM with ICL capabilities, enabling adaptation to unseen tasks. Experiments demonstrate that ICT improves the performance of state-of-the-art TSFMs by approximately 11.4% on unseen tasks without requiring fine-tuning.

Metadata

arXiv ID: 2602.20307
Provider: ARXIV
Primary Category: cs.LG
Published: 2026-02-23
Fetched: 2026-02-25 06:05

Related papers

Raw Data (Debug)
{
  "raw_xml": "<entry>\n    <id>http://arxiv.org/abs/2602.20307v1</id>\n    <title>In-context Pre-trained Time-Series Foundation Models adapt to Unseen Tasks</title>\n    <updated>2026-02-23T19:48:47Z</updated>\n    <link href='https://arxiv.org/abs/2602.20307v1' rel='alternate' type='text/html'/>\n    <link href='https://arxiv.org/pdf/2602.20307v1' rel='related' title='pdf' type='application/pdf'/>\n    <summary>Time-series foundation models (TSFMs) have demonstrated strong generalization capabilities across diverse datasets and tasks. However, existing foundation models are typically pre-trained to enhance performance on specific tasks and often struggle to generalize to unseen tasks without fine-tuning. To address this limitation, we propose augmenting TSFMs with In-Context Learning (ICL) capabilities, enabling them to perform test-time inference by dynamically adapting to input-output relationships provided within the context. Our framework, In-Context Time-series Pre-training (ICTP), restructures the original pre-training data to equip the backbone TSFM with ICL capabilities, enabling adaptation to unseen tasks. Experiments demonstrate that ICT improves the performance of state-of-the-art TSFMs by approximately 11.4% on unseen tasks without requiring fine-tuning.</summary>\n    <category scheme='http://arxiv.org/schemas/atom' term='cs.LG'/>\n    <published>2026-02-23T19:48:47Z</published>\n    <arxiv:primary_category term='cs.LG'/>\n    <author>\n      <name>Shangqing Xu</name>\n    </author>\n    <author>\n      <name>Harshavardhan Kamarthi</name>\n    </author>\n    <author>\n      <name>Haoxin Liu</name>\n    </author>\n    <author>\n      <name>B. Aditya Prakash</name>\n    </author>\n  </entry>"
}